regex solution to a sas-l problem.

gistfile1.txt

data have;
  retain prx ;
input path : $800.;
  if _n_ = 1 then do ;
    prx = prxparse("(Sect1.2.3b|Sect1.1.a|Sect2.3.b|Sect3.3.2c)") ;
  end ;
  flg = (prxmatch(prx, path) > 0) ;
  if flg then do ;
    call prxsubstr(prx, path, pos, len) ;
    excerpt = substr(path, pos, len) ;
  end ;
  drop pos len ;
datalines;
Garbage.Text.1/Garbage.Text.2/Sect1.2.3b/Garbage.Text.3
Garbage.Text.1/Sect1.1.a/Garbage.Text.2/Garbage.Text.4/Garbage.Text.5
Garbage.Text.1/Garbage.Text.2
Sect2.3.b/Garbage.Text.1
Garbage.Text.1/Garbage.Text.2/Garbage.Text.4/Garbage.Text.5/Sect3.3.2c
Garbage.Text.1/Garbage.Text.2/Garbage.Text.4/Garbage.Text.5
;
run;